Examining avian diversity with relation to snag density in wetlands ...
Transcript of Examining avian diversity with relation to snag density in wetlands ...
Examining avian diversity with relation to snag density in wetlands and woodlands in the Great
Lakes Region
BIOS 35502: Practicum in Field Biology
Kate E. Augustine
Advisor: Michael Chips
2011
Augustine 2
Abstract
Snags (standing dead trees) are important in temperate ecosystems because they provide
nesting and foraging habitat for cavity-nesting birds. Cavities excavated by these birds are
subsequently used by many other wildlife species, including secondary-cavity nesting birds.
Therefore maintaining natural snag density is important in order to maintain the natural diversity
of forest ecosystems. This study focuses on the relationship between avian diversity and snag
abundance in wetlands and unmanaged, second-growth mixed deciduous and coniferous forest
stands at the University of Notre Dame Environmental Research Center. Point count call surveys
analyzing avian species richness and abundance were undertaken during the period of 2 June and
12 July, 2011. Vegetation surveys were conducted within a 30 m radius around observation
points counting the number of snags and live trees. Tree type (i.e. coniferous or deciduous) and
snag dbh were also recorded. Species diversity did not differ between habitat type (p=0.1864),
but analysis of variance suggests that snag density does have an effect on species diversity within
sites although this relationship is not supported by regression analysis between number of snags
per plot area and species diversity (p=0.0432). However, a relationship does exist between the
proportion of coniferous snags and avian species diversity (p=0.0821). Therefore, future
management decisions pertaining to timber harvesting and fire suppression practices should
consider the use of coniferous snags by birds. Understanding the influence of snag density on
bird communities is fundamental to developing harvest methods that ensure the conservation of
forest birds.
Introduction
The management of forests for the timber industry and fire suppression drastically
changes forest dynamics, especially the abundance of coarse woody debris (CWD) within a
forest. CWD plays numerous roles in the functioning of a healthy forest and includes all dead
wood, fallen logs, branches, large roots, and snags (standing dead trees). CWD provides an input
of nutrients back into soils, long-term carbon storage, and essential habitat for various small
mammals, amphibians, arthropods, and birds (Harmon et al. 1986). Therefore it is critical that
CWD density is retained at levels found in pristine, unmanaged environments in order to
maintain the natural diversity of forest ecosystems.
Cavity-nesting birds are important components of the avifauna in forests of the Great
Lakes region, and primary cavity-nesting species are of special importance because they
excavate cavities for nesting which are subsequently used by many other species of wildlife
Augustine 3
including secondary-cavity nesting bird species. Changes in number and characteristics of snags
may affect the abundance of cavity-nesting birds because snags provide sites for nesting,
roosting, and foraging (Mannan et al. 1980, Weikel and Hayes 1999). Legacy snags (large
diameter snags from the previous stand) in young forests are especially important resources for
cavity-nesting birds providing both nesting habitats (Mannan et al. 1980, Lundquist and Mariani
1991) and foraging substrates (Weikel and Hayes 1999). Understanding the influence of snag
density on bird communities is fundamental to developing harvest methods that ensure the
conservation of forest birds.
This study focuses on the relationship between avian diversity and snag abundance in
wetlands and unmanaged, second-growth mixed deciduous and coniferous forest stands at the
University of Notre Dame Environmental Research Center (UNDERC) located in the Upper
Peninsula of Michigan. I predicted that snag densities would differ within and between habitat
types and that bird diversity would vary accordingly such that the most diverse sites would be
areas with high snag densities. Because other attributes of vegetation structure and composition
are also known to be key factors in determining habitat selection by birds (MacArthur and
MacArthur 1961), other factors such as canopy cover, proportion of live and dead trees, and the
proportion of deciduous and coniferous trees were also analyzed in this study to help further
explain observed patterns of bird diversity on the UNDERC property.
Materials and Methods
Site selection
Twelve sites were haphazardly selected at the UNDERC property to represent woodland
and wetland habitats with varying snag densities. Upon first observation, three high density snag
sites and three low snag density sites were selected for each habitat type for a total of six
Augustine 4
woodland and six wetland sites (Figure 1). The chosen sites were at least 250-m apart to
guarantee that overlap of the point count areas did not occur. Characteristics of each site are
described in Table 1. The woodland sites included both sugar maple woodlands and mixed
conifer and deciduous stands. The six wetland sites included both bogs and fens with and without
open water.
Point count call surveys
Avian richness and abundance were investigated through point count call surveys during
the period of 2 June to 12 July, 2011 Birds were not surveyed in extreme weather, when wind or
rain interfered with the audibility of bird calls, or when fog or rain impaired visibility. Surveys
were conducted once each day within 4 hours after sunrise. This procedure yielded a total of 36
bird surveys: 9 for each habitat type and snag density, and 3 replicate surveys for each individual
site.
The call surveys were conducted according to a standard point count method utilized in
bird studies by the Breeding Biology Research and Monitoring Database at the University of
Montana (Martin et al. 2007). Point counts involved an observer remaining at one location and
recording all the birds seen or heard inside and outside of a 30 m radius circle within a 30 minute
time period. Observation points were flagged so that point counts were taken from the same
location within each plot. With the exception of a few wetland sites, each point was situated at
least 30 m within the specific habitat. The species were written in the order of auditory
observation and the number of individuals for each species was also recorded. Visual encounters
were used to confirm call identification when species were within view from the observation
point. To include only species actively using observation sites, flyovers were not included unless
Augustine 5
they landed within the plot area. A digital voice recorder was also utilized during each survey to
record unknown calls which were later analyzed and identified to species.
Vegetation surveys
Tree trunks were counted within a 30 m radius surrounding each bird count point to
quantify the proportion of tree types (i.e. coniferous or deciduous, live or snag) within each plot.
Snags were identified as standing dead trees lacking green foliage. Snag diameter at breast height
(dbh) was measured and used to calculate snag basal area. For both live trees and snags, I
defined that each individual tree should stand at least 1.65m tall at an estimated angle >45
degrees, and >8cm dbh. Canopy cover was also calculated using a densiometer.
Statistical analysis
Jaccard’s similarity index was used in a cluster analysis to compare similarities in species
composition by site. The bird species diversity at each site was calculated using the Shannon-
Weiner index of diversity. Shannon-Weiner is calculated as H = ∑ pi (ln pi) where pi is the
proportion of individuals which belong to the ith
species. This index provides a quantitative
measure of the diversity of species in the total population at each particular site. This measure is
conservative because abundance data was only noted when multiple birds of the same species
were either within sight or calling back and forth.
Analysis of variance on number of snags was performed in Systat 13.0 (SPSS, INC.,
2010) to determine if any significant differences occurred between habitat types. Regression
analyses were performed to directly assess the relation between snag density and bird diversity,
as well as the relationship between snag number and bird diversity. Regressions analyzing the
relationship between bird diversity with snag basal area, cover, proportion of live trees,
Augustine 6
proportion of snags, proportion of deciduous and coniferous live trees, and the proportion of
deciduous and coniferous snags were also performed. All data sets were tested for normality in
Systat, and proportion data was transformed using arcsine square root transformation methods.
Results
A total of 191 birds, composed of 37 species, were observed during the course of this
study (Appendix 1). The most common species observed was the ovenbird, followed by the
yellow-bellied sapsucker. The ovenbird, yellow-bellied sapsucker, and blue jay were all present
at the greatest number of sites (8), followed by the Nashville warbler (7). The mean number of
individuals per survey at each site varied from 3 to 8.33, with a mean 5.31. Species richness per
site ranged from 6 to 15 with a mean of 10.41. Bird species diversity ranged from 1.677 to 2.626,
with an mean of 2.184 (Table 2).
Cluster analysis indicates that generally woodland locations were similar in species
composition, and wetland locations were also similar in species composition (Table 3, Figure 2).
Total snag density recorded as number of snags within each plot did not differ between wetland
and woodland habitats (F1,10 =0.1187, p=0.7376, Figure 2). Mean snag density within wetland
habitats was 24 snags, and within woodland habitats it was 28.33 snags. Total number of snags
did differ between high and low density snag sites (F1,10=16.775, p=0.002, Figure 3). Mean snag
density within high density sites was 101.5 snags, and within low density snag sites it was 93
snags. Species diversity, as calculated by the Shannon-Wiener index did not differ between
habitats (F1,10 =2.0124, p=0.1864, Figure 5). Mean species diversity within wetlands was 2.32,
and within woodland mean species diversity was 2.05. Species diversity was significantly
different between high and low density snag sites (F1,10 =5.3581, p=0.0432, Figure 6). Mean
Augustine 7
species diversity within high density snag sites was 2.38, and within low density snag sites it was
1.99.
Regression analysis of basal area of snags versus species diversity did not show a
relationship (F1,10 =0.003485, R2=0.0003, p=0.9541, Figure 7). Canopy cover versus species
diversity also did not show a relationship (F1,10 =0.988754, R2=0.09, p=0.3435, Figure 8). A
relationship also does not exist between number of snags and species diversity (F1,10 =3.7348,
R2=0.2719, p=0.0821, Figure 9). Species diversity also did not show relationships with the
proportion of live trees (F1,10 =2.575181, R2=0.2048, p=0.1396, Figure 10), proportion of dead
trees (F1,10 =2.575181, R2=0.2048, p=0.1396, Figure 11), proportion of live deciduous trees
(F1,10 =2.8784, R2=0.2235, p=0.1206, Figure 12), proportion of live coniferous trees (F1,10
=2.8784, R2=0.2235, p=0.1206, Figure 13), or the proportion of dead deciduous trees (F1,10
=0.1986, R2=0.0195, p=0.6653, Figure 14). However, species diversity does show a significant
relationship with dead coniferous snags (F1,10 =8.42062, R2=0.4571, p=0.0158, Figure 15).
Discussion
Although Shannon-Wiener diversity indexes did not differ by habitat (Figure 5), cluster
analysis (Figure 2) indicates that overall species composition varied between sites but species
composition among woodland sites was generally more, and wetland sites were also generally
more similar. However, two wetland sites have species composition more similar to the
woodland sites than other wetland sites, but this can be mostly explained by wetland type. Site 9
is likely more similar to woodland species composition because it is a smaller bog surrounded by
mixed coniferous and deciduous forest and all birds observed in this area visited this site from
the surrounding forest edge. Site 10 is an almost completely open bog with standing water, no
snags, and very few trees located within the plot area. It is therefore likely that most species
Augustine 8
observed here were simply visiting the site to take advantage of the open water from the
surrounding forest edge.
Site 12 is primarily coniferous woodland that differs from both woodland and wetland
habitats in its species composition (Figure 2). This is due to the fact that it was the sole location
that two species, including the veery and white-breasted nuthatch, were observed during this
study. White-breasted nuthatches likely prefer this site because they favor mature forests
(Peterson 2010) and Site 12 is located at the north end of property on the edge of the Ottawa
National Forest. The veery prefers willow and alder thickets along streams (Peterson 2010) and
such habitat did border one side of this location because the northern area of property has been
heavily modified by beaver activity.
Cavity nesting birds use a variety of decayed trees and snags for nesting, foraging, and
roosting (Mannan et al. 1980). In prior studies, population densities of cavity nesters are
positively correlated to snag density (Raphael and White 1984, Land et al. 1989, Lohr et al.
2002) and as a result, lower densities of cavity nesting birds have been identified as an important
factor causing lower total bird density in managed forest stands (Nilsson 1979). Contrary to
prior research and my original hypothesis however, this study did not find a relationship between
bird diversity and snag density, either measured by basal area (Figure 7) or number of snags
(Figure 9) although analysis of variance indicated that both snag density and calculated Shannon-
Wiener diversity indexes were significantly different between high and low density snag sites
(Figs. 4 & 6). It is not surprising that bird diversity in this region is not affected by snag basal
area. Bird species that specifically require trees with larger basal area (>100 cm dbh) are likely
not observed within my selected study sites because out of the twelve site studied, the largest
snags found were <50 cm dbh. It has been observed that certain species do favor large snags over
Augustine 9
small snags when available (Mannan et al. 1980), including the pileated woodpecker which does
occur on property and was observed within the course of this study. Therefore it would be
interesting to further study individual snag use to further analyze the relationship between bird
diversity and snag size.
The lack of apparent snag use could also be explained by prey availability and
accessibility. In addition to providing nesting sites, birds also use snags for foraging. Snags
harbor a different insect fauna than found in live trees and therefore type of insect prey
availability in snags may have discouraged many birds, particularly birds that feed on insects
found on foliage, from extensively using them (Franzeb 1978). Prey accessibility might also
have limited snag use because feeding on insects living in snags requires morphological
adaptations (Franzeb 1978).
The lack of a relationship between snag density and species diversity could also be the
result of survey techniques. It is likely that more replicates per site could result in a significant
relationship between number of snags and bird diversity, but due to a late project change and
inclement weather during both research weeks I was only able to complete the minimum three
replicates per observational site. The integrity of the first week of observations I made might also
be decreased because I had no prior experience birding before the start of this study. Realizing
this, I did attempt to repeat my initial five surveys so as to not include data that I was uncertain
of in my analyses. Certain assumptions are made when performing point count call surveys
which result in conservative observations of species diversity within sites because an observer
can only be certain more than one individual of each species is located within a site if multiple
birds were either within sight or calling back and forth. Therefore mist netting techniques would
Augustine 10
be able to obtain more accurate representations of abundance and bird species diversity within
sites, enabling birds that do not call or sing to also be observed.
Avian diversity has also been found to correlate with other environmental factors
including amount of foliage and vegetation structure (MacArthur and MacArthur 1961), and tree
selection depends on food abundance, availability, and quality (Franzeb 1978). Foliage may be
important for birds in that it protects them from predators and inclement weather conditions and
shelters nest sites so that birds do not prefer using tree species or habitats associated with open
vegetation (Franzreb 1978). I observed the foliage cover differed between woodland and wetland
environments, with woodlands having generally more dense canopies than wetlands but I did not
find a significant difference in species diversity between habitats nor a relationship between
cover and species diversity (Figure 8). Although I did not find a significant relationship, it would
be interesting to further study the relationship between foliage stratification and individual
species use of foliage layers within a site, as well as tree species use comparing foliage volume
between tree species.
Tree species selection has also been found to influence bird diversity, providing a variety
of foraging habitats (Franzreb 1978, Wiekel and Hayes 1999). Deciduous trees are an important
foraging substrate for species that forage primarily on live foliage (Franzreb 1978), and Schimpf
and MacMahon (1985) found that arthropod density was higher in canopies of deciduous aspen
forests than in canopies of coniferous forests. Wiekel and Hayes (1999) also suggest that the
smaller passerine species, which include warblers and goldfinches, favor conifers because large
leaf size of deciduous trees makes it difficult to perch on a branch and reach the outer portion of
the leaves which harbor insects. Because most bird species observed within the course of this
study are insectivorous, I therefore predicted that bird species diversity might vary not only with
Augustine 11
the proportion of snags, but that bird diversity might also show a relationship with varying
proportions of deciduous and coniferous trees. It is interesting that the relationship between
proportion of coniferous snags and species diversity was the only significant relationship found
in this study, such that diversity appears to increase with increasing proportion of coniferous
snags.
The significant relationship between coniferous snags and bird diversity could possibly
be explained by the presence and quantity of dead branches found on coniferous snags which
may influence foraging activity by cavity-nesting birds (Wiekel and Hayes 1999). Dead
branches may be preferred foraging habitats because they have more sloughing bark which
provides habitat for many arthropods including spiders, and therefore trees with more dead
branches provide greater prey densities than snags with fewer dead branches (Holmes and
Robinson 1981). Although number of branches was not recorded in this study, it was observed
that coniferous snags tended to have a greater number of branches than deciduous snags. It
should be noted that coniferous trees were the most common tree type observed within study
sites and that the relationship of dead coniferous snags and species diversity might therefore be
affected by availability. My study did not directly address selection of stand types or
microhabitats within a stand for foraging by cavity-nesting birds, but my results suggest that
selection may be occurring at these scales and therefore further study on tree species preferences
is necessary.
In summary, this study did not find a relationship between avian species diversity and
snag density on the UNDERC property. However, avian species diversity was found to be related
to increasing proportion of coniferous snags. Therefore, future management decisions pertaining
to timber harvesting and fire suppression practices should consider the use of coniferous snags
Augustine 12
by birds. The majority of coniferous snags should not be harvested because they serve several
significant functions such as providing nesting and foraging sites for numerous cavity-nesting
species. In order to understand factors affecting avian species diversity within the UNDERC
property and ensure proper conservation of forest birds, further study is necessary to determine
factors affecting stand type selection by cavity-nesting species.
Acknowledgements
I especially thank my mentor Michael Chips for his guidance, patience, and assistance
throughout all phases of this study. I thank Dr. Michael Cramer, Heidi Mahon, Matt Igleski, and
Shayna Sura for their assistance with ideas and statistical analysis throughout the summer at
UNDERC. Special thanks to Kate Kirbie for assistance in vegetation measurements. Additional
thanks to the Bernard J. Hank Family for their generous funding allowing this research to be
possible.
Augustine 13
References
Franzred, K.E. 1978. Tree species used by birds in logged and unlogged mixed-coniferous forests. The
Wilson Bulleting 90: 221-238.
Harmon, M.E., J.F. Franklin, F.J. Swanson, P. Sollins, S.V. Gregory, J.D. Lattin, N.H. Anderson, S.P.
Cline, N.G. Aumen, J.R. Sedell, G.W. Lienkaemper, K. Cromack Jr., and K.W. Cummins. 1986.
Ecology of coarse woody debris in temperate ecosystems. Advances in Ecological Research 15:
133-302.
Holmes, R.T., and S.K. Robinson. 1981. Tree species preferences of foraging insectivorous birds in a
northern hardwoods forest. Oecologia 48: 31-35.
Land, D., W.R. Marion, and T.E. O’Meara. 1989. Snag availability and cavity nesting birds in slash pine
plantations. Journal of Wildlife Management 53: 1165-1171.
Lohr, S.M., S.A. Gauthreaux, and J.C. Kilgo. Importance of coarse woody debris to avian communities in
loblolly pine forests. Conservation Biology 16: 767-777.
Lundquist, R.W. and J.M. Mariani. 1991. Nesting habitat and abundance of snag-dependent birds in the
southern Washington Cascade Range in Wildlife and vegetation of unmanaged Douglas-fir
forests. USDA Forest Service General Technical Report 221-240.
MacArthur, R.H., and J.W. MacArthur. 1961. On bird species diversity. Ecology 42: 594-598.
Martin, T.E., C. Paine, C.J. Conway, W.M. Hochachka, P. Allen, and W. Jenkins. 1997. The Breeding
Biology Research & Monitoring Database. Retrieved 26 May 2011, from Montana Coop.
Wildlife Research Unit: http://www.umt.edu/bbird/protocol/pointcnt.aspx
Mannan, R.W., E.C. Meslow, and H.M. Wight. Use of snags by birds in Douglas-fir forests, Western
Oregon. Journal of Wildlife Management 44: 787-797.
Nilsson, S.G. 1979. Effect of forest management on the breeding bird community in southern Sweden.
Biological Conservation 16: 135-143.
Peterson, R.T. 2010. Peterson field guide to birds of eastern and central North America, 6th edition.
Boston: Houghton Mifflin Harcourt Publishing Company.
Schimpf, D.J. and J.A. MacMahon. 1985. Insect communities and faunas of a Rocky Mountain subalpine
sere. Western North American Naturalist 45: 37-60.
SPSS, Inc. 2010, SYSTAT version 13.0, Chicago, IL
Raphael, M.G., and M. White. 1984. Use of snags by cavity-nesting birds in the Sierra Nevada. Wildlife
Monographs 86: 1-66.
Weikel, J.M., and J.P. Hayes. 1999. The foraging ecology of cavity-nesting birds in young forests of the
Northern coast range of Oregon. The Condor 101: 58-66.
Augustine 14
Table 1. Characteristics of study sites at the UNDERC property. Sites were surveyed during the period
of June 2 to July 12, 2011.
Type Snag density Description
Site 1 Wetland High Bog, emergent vegetation consisting primarily of tall grasses
and speckled alder bushes. Lacking floating vegetation mat.
Site 2 Woodland High Mixed coniferous and deciduous woodland.
Site 3 Wetland High Open water bog with some emergent vegetation and multiple
smaller floating vegetation mats.
Site 4 Woodland Low Deciduous woodland dominated primarily by aspens.
Site 5 Wetland Low Bog, with some open water and a large floating vegetation mat
formed from sphagnum. No snags.
Site 6 Woodland High Mixed coniferous and deciduous woodland, high density
course woody debris.
Site 7 Wetland High Tamarac bog associated with sphagnum mat, no open water.
Site 8 Woodland Low Sugar maple woodland.
Site 9 Wetland Low Small Tamarac bog dominated by sphagnum, lacking open
water.
Site 10 Wetland Low Open water bog associated with tall emergent vegetation,
lacking floating vegetation. No snags.
Site 11 Woodland Low Mixed woodland lacking dense understory and dominated
primarily by spruce. Forest floor covered in sphagnum.
Site 12 Woodland High Mixed coniferous and deciduous woodland dominated
primarily by pines.
Augustine 15
Table 2. Shannon-Wiener diversity indexes for species composition within sites surveyed
during the summer of 2011 at UNDERC. Site codes are as follows: Habitat type (W=woodland,
B=wetland), snag density classification (H=high, L=low), and replicate number.
Site # Code Species (H')
Site 1 BH1 2.553454532
Site 2 WH1 2.394700406
Site 3 BH2 2.626164786
Site 4 WL1 1.846220219
Site 5 BL1 2.553237003
Site 6 WH2 1.859428815
Site 7 BH3 2.491493722
Site 8 WL2 1.676987774
Site 9 BL2 1.907283999
Site 10 BL3 1.798652206
Site 11 WL3 2.139085895
Site 12 WH3 2.36938212
Table 3. Jaccard’s distance calculated from Jaccard’s index in Cluster analysis. Sites were
surveyed at UNDERC during the period of June 2 to July 12, 2011.
Clusters Joining at Distance No. of Members
BL2 WH1 0.571429 2
BH3 BH1 0.578947 2
BH3 BL1 0.600000 3
WL2 BL2 0.600000 3
WL3 WH2 0.615385 2
WL2 WL3 0.615385 5
WL1 WL2 0.636364 6
BH2 BH3 0.666667 4
BL3 WL1 0.692308 7
BH2 BL3 0.700000 11
WH3 BH2 0.700000 12
Augustine 16
Figure 1. Study sites on UNDERC property surveyed during the summer of 2011. Black squares are high
density snag woodlands, and white squares are low density snag woodlands. Black circles sites are high
density snag wetlands, and white circles sites are low density snag wetlands.
Figure 2. Cluster analysis based on distances calculated from Jaccard’s index. Woodland sites are
generally similar in species composition, and wetland sites are also similar with the exception of Site 9
and Site 10. Site 12 had two species observed at no other location on property. Sites were surveyed at
UNDERC during the summer of 2011.
Woodland, high snag
Woodland, low snag
Wetland, high snag
Wetland, low snag
Augustine 17
Figure 3. Mean number of snags between wetland and woodland sites (standard error shown). Sites were
surveyed at UNDERC during the summer of 2011.
Figure 4. Mean number of snags between “high” and “low” density snag sites (standard error shown).
Sites were surveyed at UNDERC during the summer of 2011.Large standard error likely results from the
difficulty of sampling in wetland environments.
24
28.33
0
5
10
15
20
25
30
35
40
1
Nu
mb
er
of
snag
s p
er
plo
t
Wetlands Woodlands
101.5 93
0
20
40
60
80
100
120
140
160
180
1
Nu
mb
er
of
snag
s p
er
plo
t ar
ea
High Low
Augustine 18
Figure 5. Mean Shannon-Wiener diversity between wetland and woodland sites (standard error shown).
Sites were surveyed at UNDERC during the summer of 2011.
Figure 6. Mean Shannon-Wiener diversity between “high” and “low” density snag sites (standard error
shown). Sites were surveyed at UNDERC during the summer of 2011.
2.32
2.05
0
0.5
1
1.5
2
2.5
3
1
Spe
cie
s d
ive
rsit
y (H
')
Wetlands Woodlands
2.38
1.99
0
0.5
1
1.5
2
2.5
3
1
Spe
cie
s d
ive
rsit
y (H
')
High Low
Augustine 19
Figure 7. Relationship between basal area of snags (m
2) and Shannon-Wiener species diversity index
(H’) (p=0.9541). Sites were surveyed at UNDERC during the summer of 2011.
Figure 8. Relationship between canopy cover and Shannon-Wiener species diversity index (H’) (p=0
.3435). Sites were surveyed at UNDERC during the summer of 2011.
y = 0.0148x + 2.1768 R² = 0.0003
0
0.5
1
1.5
2
2.5
3
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Spe
cie
s D
ive
rsit
y (H
')
Basal area of snags (m2)
y = -0.0036x + 2.3734 R² = 0.09
0
0.5
1
1.5
2
2.5
3
0 20 40 60 80 100
Spe
cie
s D
ive
rsit
y (H
')
Canopy cover (%)
Augustine 20
Figure 9. Relationship between snag number found in the plot area (30 m radius) and Shannon-Wiener
species diversity index (H’) (p=0.0821). Sites were surveyed at UNDERC during the summer of 2011.
Figure 10. Relationship between proportion of live trees and Shannon-Wiener species diversity index
(H’) (p=0.1346). Sites were surveyed at UNDERC during the summer of 2011.
y = 0.0087x + 1.9563 R² = 0.2719
0
0.5
1
1.5
2
2.5
3
0 10 20 30 40 50 60 70
Spe
cie
s D
ive
rsit
y (H
')
Number of snags
y = -0.6919x + 3.0352 R² = 0.2048
0
0.5
1
1.5
2
2.5
3
0 0.5 1 1.5 2
Spe
cie
s D
ive
rsit
y (H
')
Proportion of live trees
Augustine 21
Figure 11. Relationship between proportion of dead trees and Shannon-Wiener species diversity index
(H’) (p=0.1396). Sites were surveyed at UNDERC during the summer of 2011.
Figure 12. Relationship between proportion of live deciduous trees and Shannon-Wiener species
diversity index (H’) (p=0.1206). Sites were surveyed at UNDERC during the summer of 2011.
y = -0.4509x + 2.367 R² = 0.2235
0
0.5
1
1.5
2
2.5
3
0 0.2 0.4 0.6 0.8 1 1.2
Spe
cie
s D
ive
rsit
y (H
')
Proportion of live deciduous trees
y = 0.6919x + 1.9484 R² = 0.2048
0
0.5
1
1.5
2
2.5
3
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Spe
cie
s D
ive
rsit
y (H
')
Proportion of dead trees
Augustine 22
Figure 13. Relationship between proportion of live coniferous trees and Shannon-Wiener species
diversity index (H’) (p=0.1206). Sites were surveyed at UNDERC during the summer of 2011.
Figure 14. Relationship between proportion of dead deciduous trees and Shannon-Wiener species
diversity index (H’) (p=0.6653). Sites were surveyed at UNDERC during the summer of 2011.
y = 0.4509x + 1.9162 R² = 0.2235
0
0.5
1
1.5
2
2.5
3
0 0.2 0.4 0.6 0.8 1 1.2
Spe
cie
s D
ive
rsit
y (H
')
Proportion of live coniferous trees
y = -0.9255x + 2.2351 R² = 0.0195
0
0.5
1
1.5
2
2.5
3
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Spe
cie
s D
ive
rsit
y (H
')
Proportion of dead deciduous trees
Augustine 23
Figure 15. Relationship between proportion of dead coniferous trees and Shannon-Wiener species
diversity index (H’) (p=0.0158). Sites were surveyed at UNDERC during the summer of 2011.
y = 0.9461x + 1.9783 R² = 0.4571
0
0.5
1
1.5
2
2.5
3
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Spe
cie
s D
ive
rsit
y (H
')
Proportion of dead coniferous trees
Augustine 24
APPENDIX
Table A-1. Abundance and distribution of observed bird species at the UNDERC property between 2
June and 12 July, 2011. Codes follow English names prepared by the Institute for Bird Populations.
Species
Total No. of
Individuals Observed
Total No. of Sites
Where Present
NAWA 9 7
NOPA 7 4
YWAR 5 5
CMWA 7 5
YRWA 3 3
BTNW 5 3
BAWW 5 4
AMRE 3 2
OVEN 22 8
COYE 6 3
RBGR 1 1
CHSP 5 4
SOSP 6 4
SWSP 11 5
WTSP 10 4
RWBL 3 3
COGR 5 2
AMGO 2 2
RTHU 3 2
BEKI 3 2
YBSA 14 8
NOFL 1 1
PIWO 3 3
OSFL 1 1
ALFL 2 2
LEFL 2 2
EAKI 5 3
WAVI 1 1
REVI 9 6
BLJA 8 8
VEER 1 1
HETH 5 4
AMRO 3 3
WOTH 2 1
CEDW 10 5
WBNU 1 1
HAWO 2 2
Augustine 25
Table A-2. Count data on the number of live trees and snags by tree type within a 30 m radius plot area
from each observation site on the UNDERC property. Sites were surveyed at UNDERC during the
summer of 2011.
Live trees Snags
Site Coniferous Deciduous Coniferous Deciduous
1 35 46 10 16
2 111 119 40 9
3 52 33 44 9
4 0 159 0 25
5 34 1 0 0
6 65 125 8 23
7 78 0 63 0
8 0 159 0 12
9 22 21 0 2
10 19 0 0 0
11 483 20 21 2
12 268 53 13 17